As a kind of statistical method, the technique of Hidden Markov Model （HMM） is widely used for speech recognition. In order to train the HMM to be more effective with much less amount of data, the Subspace Distribut...As a kind of statistical method, the technique of Hidden Markov Model （HMM） is widely used for speech recognition. In order to train the HMM to be more effective with much less amount of data, the Subspace Distribution Clustering Hidden Markov Model （SDCHMM）, derived from the Continuous Density Hidden Markov Model （CDHMM）, is introduced. With parameter tying, a new method to train SDCHMMs is described. Compared with the conventional training method, an SDCHMM recognizer trained by means of the new method achieves higher accuracy and speed. Experiment results show that the SDCHMM recognizer outperforms the CDHMM recognizer on speech recognition of Chinese digits.展开更多
If K ∩ AlgL is weak. dense in AlgL, where K is the set of all compactoperators in B(H), is completely distributive? In this note, we prove that there is a reflexivesubspace lattice L on some Hilbert space, which sati...If K ∩ AlgL is weak. dense in AlgL, where K is the set of all compactoperators in B(H), is completely distributive? In this note, we prove that there is a reflexivesubspace lattice L on some Hilbert space, which satisfies the following conditions: (a) F(AlgL) isdense in AlgL in the ultrastrong operator topology, where F(AlgL) is the set of all finite rankoperators in AlgL; (b) L isnt a completely distributive lattice. The subspace lattices that satisfythe above conditions form a large class of lattices. As a special case of the result, it easy to seethat the answer to Problem 7 is negative.展开更多
基金Supported by the National Natural Science Foundation of China (No.60172048)
文摘As a kind of statistical method, the technique of Hidden Markov Model （HMM） is widely used for speech recognition. In order to train the HMM to be more effective with much less amount of data, the Subspace Distribution Clustering Hidden Markov Model （SDCHMM）, derived from the Continuous Density Hidden Markov Model （CDHMM）, is introduced. With parameter tying, a new method to train SDCHMMs is described. Compared with the conventional training method, an SDCHMM recognizer trained by means of the new method achieves higher accuracy and speed. Experiment results show that the SDCHMM recognizer outperforms the CDHMM recognizer on speech recognition of Chinese digits.
文摘If K ∩ AlgL is weak. dense in AlgL, where K is the set of all compactoperators in B(H), is completely distributive? In this note, we prove that there is a reflexivesubspace lattice L on some Hilbert space, which satisfies the following conditions: (a) F(AlgL) isdense in AlgL in the ultrastrong operator topology, where F(AlgL) is the set of all finite rankoperators in AlgL; (b) L isnt a completely distributive lattice. The subspace lattices that satisfythe above conditions form a large class of lattices. As a special case of the result, it easy to seethat the answer to Problem 7 is negative.
